SERIES

PHYSIOLOGY IN RESPIRATORY MEDICINE

Dead space: the physiology of wasted

ventilation
H. Thomas Robertson

Number 9 in the series “Physiology in respiratory medicine”

Edited by R. Naeije, D. Chemla, A. Vonk-Noordegraaf and A.T. Dinh-Xuan

Affiliation:
Division of Pulmonary and Critical Care Medicine, University of Washington, Seattle, WA, USA.

Correspondence:
H. Thomas Robertson, Division of Pulmonary and Critical Care Medicine, Depts of Medicine and Physiology
and Biophysics, University of Washington, Box 356522 Seattle, WA, 98195-6522, USA.
E-mail: tomrobt@uw.edu

ABSTRACT An elevated physiological dead space, calculated from measurements of arterial CO2 and
mixed expired CO2, has proven to be a useful clinical marker of prognosis both for patients with acute
respiratory distress syndrome and for patients with severe heart failure. Although a frequently cited
explanation for an elevated dead space measurement has been the development of alveolar regions
receiving no perfusion, evidence for this mechanism is lacking in both of these disease settings. For the
range of physiological abnormalities associated with an increased physiological dead space measurement,
increased alveolar ventilation/perfusion ratio (V′A/Q′) heterogeneity has been the most important
pathophysiological mechanism. Depending on the disease condition, additional mechanisms that can
contribute to an elevated physiological dead space measurement include shunt, a substantial increase in
overall V′A/Q′ ratio, diffusion impairment, and ventilation delivered to unperfused alveolar spaces.

@ERSpublications
A review of current understanding of factors accounting for abnormal physiological dead space
measurements in disease http://ow.ly/Dnyw1

Previous articles in this series: No. 1: Naeije R, Vachiery J-L, Yerly P, et al. The transpulmonary pressure gradient for
the diagnosis of pulmonary vascular diseases. Eur Respir J 2013; 41: 217–223. No. 2: Hughes JMB, van der Lee I. The
TL,NO/TL,CO ratio in pulmonary function test interpretation. Eur Respir J 2013; 41: 453–461. No. 3: Vonk-Noordegraaf A,
Westerhof N. Describing right ventricular function. Eur Respir J 2013; 41: 1419–1423. No. 4: Hamzaoui O, Monnet X,
Teboul J-L. Pulsus paradoxus. Eur Respir J 2013; 42: 1696–1705. No. 5: Prisk GK. Microgravity and the respiratory
system. Eur Respir J 2014; 43: 1459–1471. No. 6: Dempsey JA, Smith CA. Pathophysiology of human ventilatory control.
Eur Respir J 2014; 44: 495–512. No. 7: Petersson J, Glenny RW. Gas exchange and ventilation–perfusion relationships in
the lung. Eur Respir J 2014; 44: 1023–1041. No. 8: Wagner PD. The physiological basis of pulmonary gas exchange:
implications for clinical interpretation of arterial blood gases. Eur Respir J 2015; 45: 227–243.
Received: July 27 2014 | Accepted after revision: Sept 29 2014 | First published online: Nov 13 2014
Conflict of interest: None declared.
Copyright ©ERS 2015

1704 Eur Respir J 2015; 45: 1704–1716 | DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

Introduction
Recent recognition of the prognostic relevance of measurements of physiological dead space for patients
with the acute respiratory distress syndrome (ARDS) has brought new attention to a simple gas exchange
calculation described over 120 years ago. The current calculation of physiological dead space, utilising
measurements of arterial CO2 tension (PaCO2) and mixed expired CO2 tension (PECO2), was initially
thought to include an anatomical dead space, representing the fraction of ventilation advancing no further
than the conducting airways, and an alveolar dead space, representing the fraction of ventilation delivered
to alveolar surfaces receiving no pulmonary artery perfusion. With the subsequent development of a simple
measurement of anatomical dead space, it became apparent that the alveolar dead space component was
substantially increased in a range of pulmonary diseases, and that its original interpretation as ventilation
delivered to unperfused alveolar surfaces was not adequate to explain the pathophysiology of most disease
conditions. This review will cover the findings in gas exchange pathophysiology that shape our current
understanding of the factors that influence physiological dead space measurements, with a primary focus
on dead space measurements acquired in patients with respiratory or cardiac diseases.

Bohr dead space

Because of the tidal nature of ventilation, every exhaled breath contains a fraction of the inspired gas that
does not participate in gas exchange. In 1891, the Danish respiratory physiologist Christian Bohr
introduced his calculation to represent the volume of gas within the conducting airways that constituted
the respiratory dead space [1]. The Bohr model conceptually divided the volume of the exhaled breath
(VT) into two compartments, with the first (VA) representing the fraction of exhaled breath participating
in gas exchange, and the second (VD) representing the fraction of exhaled breath penetrating no further
than the conducting airways, the respiratory dead space:

VT ¼ VA þ VD

The Bohr calculation utilised two measurements of exhaled CO2: the fractional CO2 concentration in the
total mixed exhaled breath (FECO2), and an estimate of mean alveolar CO2 concentration (FACO2) based on
a sample of gas collected late in exhalation. The mass balance calculation using CO2 assumes the dead
space contributes no CO2 to the exhaled breath (FDCO2 = 0):

V T ¼ F ECO2 ¼ V A  F ACO2 þ V D  F DCO2

Substituting (VT−VD) for VA yields:

V T  F ECO2 ¼ (V T  V D)  F ACO2

Rearranging the terms describes VD/VT, the fraction of ventilation not contributing to CO2 exchange:

V D=V T ¼ (F ACO2  F ECO2 )=F ACO2

The Bohr estimates of the volume of respiratory dead space correlated reasonably well with anatomical
measurements of the respiratory pathway, and represented a noninvasive means of measuring the extent of
ventilation inefficiency attributable to the constraints of tidal ventilation.
In the decades following the introduction of the Bohr dead space estimate, two factors that interfered with
the accuracy of the measurement as an index of conducting airways volume became apparent. The
development of a method to sample multiple within-breath samples of CO2 [2] demonstrated modest
variability in partial pressures within an exhaled breath in normal subjects, leading to the recognition that
a gas sample captured within a breath would not necessarily reflect the true mean alveolar gas
composition. While that effect was minor in normal subjects, it became a substantial source of variability
in patients with significant underlying lung disease. Furthermore, Bohr calculations performed on such
patients revealed increases in dead space fraction that could not be ascribed to conducting airway volume.
While the latter finding was an insight that drove subsequent investigations of gas exchange
pathophysiology, problems with identifying a simple method to identify mean alveolar CO2 tension
(PACO2) in patients with lung disease led to the later abandonment of the traditional Bohr measurement.
Current airway CO2 monitoring systems for intubated patients provide accurate measurements of end-tidal
CO2 concentrations that are acceptable estimates of alveolar CO2 in normal subjects, and reflect at least
directional changes in alveolar CO2 in patients with underlying lung disease, but cannot provide an accurate

DOI: 10.1183/09031936.00137614 1705

PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

measurement of Bohr dead space in a diseased lung because the end-tidal CO2 does not reflect the mean
alveolar CO2. One current application of the traditional Bohr measurement that substitutes end-tidal CO2
measurements for mean alveolar CO2 is seen with commercial exercise testing systems that calculate a Bohr
dead space during exercise. At rest, CO2 concentrations measured within a single exhalation increase
modestly, a consequence of a constant mixed venous CO2 delivery and the intermittent inspiration of fresh
gas into the alveoli. However, during heavy exercise, when mixed venous CO2 tension (PCO2) may exceed 60
mmHg, exhaled CO2 concentrations increase substantially within the course of each exhaled breath. Hence
during heavy exercise, although mean alveolar CO2 concentrations within an exhaled breath still
approximate arterial values, the end-tidal values are substantially higher than the mean alveolar values. JONES
et al. [3] documented end-tidal PCO2 measurements exceeding PaCO2 measurements by 4–6 mmHg during
heavy exercise. Hence during heavy exercise the Bohr estimate of dead space utilising measurements of
end-tidal CO2 yields an inappropriately high value because the end-tidal CO2 consistently exceeds the mean
alveolar (and arterial) PCO2. A recent method to identify an accurate mean alveolar CO2 from an expiratory
capnogram has been validated in experimental animals, and could be incorporated in commercial ventilator
systems to calculate a Bohr dead space [4]. However for critically ill patients, that calculation substantially
underestimates the more familiar Enghoff modification of the Bohr dead space, which substitutes the PaCO2
measurement for the mean PaCO2 measurement [4]. (See the physiological dead space section for
mechanisms accounting for increased arterial–alveolar PCO2 differences in abnormal lungs.)

Fowler dead space

The final fraction of each inspired breath travels no further than the conducting airways, and that
anatomical dead space volume includes the upper airways, the larynx and the volume of the
tracheo-bronchial tree extending to the acini. In the late 1940’s, with the development of a rapidly
responding nitrogen meter, Ward Fowler assembled apparatus to measure that anatomical dead space
utilising measurements of exhaled nitrogen concentrations immediately following the inspiration of a
breath of 100% oxygen [5]. Fowler’s apparatus plotted the volume of the exhaled breath following a single
inspiration of 100% oxygen against the exhaled nitrogen concentration, yielding the curve illustrated in
figure 1. The initial fraction of the exhaled breath, termed phase I, contains no nitrogen. The following
segment of the exhaled volume, termed phase II, is characterised by progressively increasing concentrations

Calv
a

b
Cinsp

Vds Valv

Vexp

FIGURE 1 Fowler’s illustration of the measurement of anatomical dead space, plotting exhaled concentration of nitrogen
(C) following an inspired breath of 100% oxygen against exhaled volume (Vexp), where Cinsp represents the inspired
nitrogen concentration (0%) and Calv represents the alveolar concentration of nitrogen. The vertical dashed line is
positioned so that a and b subsume equal areas, and the intersection of the dashed line with the exhaled volume axis
defines Vds, the anatomical dead space volume. Valv: alveolar volume. The original notation used by Fowler has been
retained. Reproduced from [5] with permission from the publisher.

1706 DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

of nitrogen prior to reaching the plateau that represents the nitrogen concentration in alveolar gas, termed
phase III. The rising slope of phase II reflects the length differences among the large airway pathways
within the lung, with shorter length conducting airways contributing alveolar gas concentrations in phase
II, and longer regions still contributing 100% oxygen from dead space gas. Fowler’s graphical method to
account for the dead space contribution from phase II was to place a straight line along phase III and adjust
a vertical line within phase II so that the two areas labelled A and B in figure 1 represent equal areas.
The exhaled volume identified by the intersection of that vertical line with the volume axis on the abscissa
identifies the dead space, a measurement he termed the physiological dead space, but is now described as
the anatomical dead space or Fowler dead space [6].
The Fowler dead space measurement is dependent on the subject size. An approximation suggested by
Fowler based on measurements in 45 subjects was that anatomical dead space in cubic centimetres roughly
equalled a subject’s ideal body weight in pounds [5]. While that anatomical dead space is usually
considered a fixed quantity, conducting airway diameter is dependent on lung volume, and when Fowler
compared measurements made at different end-inspiratory lung volumes he noted an average 100 cm3
difference in dead space measurements between the largest and smallest starting volumes. In addition,
Fowler demonstrated that the measured dead space would decrease if a 20-s breath-hold preceded the
exhalation. He postulated this effect primarily arose because of gas diffusion from alveolar regions back
into the conducting airways during the breath hold. While a subsequent study did not show significant
differences among Fowler measurements carried out with both nitrogen and helium [7], an influence of
gas diffusivity was later documented as part of a study utilising inspirations of helium (atomic weight 4)
and SF6 (molecular weight 146), demonstrating a smaller Fowler dead space with helium compared with
SF6 [8]. Gas mixing between small airways and alveolar spaces also takes place as a result of cardiogenic
movement, and this effect could also explain the progressive decrease in Fowler dead space measurements
observed after a sustained breath hold. The importance of cardiogenic motion in Fowler dead space
measurements was documented in an animal study utilising dead space measurements made before and
after cardiac arrest, with a 20% larger dead space measurement obtained after cardiac arrest [9]. A
subsequent study established those cardiogenic oscillations as more important than molecular diffusion in
reducing the dead space measured by the Fowler method [10]. Hence cardiogenic lung motion and a more
modest contribution from gas diffusion combine to decrease the measured Fowler dead space compared
with the actual volume of the respiratory pathway between the lips and the respiratory bronchioles.

Physiological dead space

In 1938, in recognition of the problem with obtaining an appropriate estimate of mean PaCO2 for
calculations using the Bohr equation, Enghoff proposed the substitution of PaCO2 for the mean PaCO2
estimate in the Bohr calculation [11]:

V D=V T ¼ (PaCO2  PECO2 )=PaCO2

Where PECO2 is the PCO2 in the total mixed exhaled breath. It is the Enghoff modification of the Bohr
dead space calculation that is in general use today, and is described as the physiological dead space, where
that term represents the sum of the anatomical dead space and the alveolar dead space, where in current
usage the alveolar dead space component is defined by the difference between the physiological dead space
and the anatomical dead space [12]. The substitution of PaCO2 for the PaCO2 avoided the difficulty of
identifying an appropriate mean PaCO2. However this modification of the Bohr equation added a new
component to the dead space estimate that is particularly relevant when the equation is utilised to describe
gas exchange in disease. The mean PaCO2 is always less than mean PaCO2, although this difference is quite
small in normal lungs. However, just as any gas exchange abnormality will increase the alveolar–arterial
O2 difference, the same statement holds for the arterial–alveolar CO2 difference. Hence the physiological
dead space, incorporating the PaCO2, will always be greater than the classic Bohr dead space, and the
physiological factors determining the arterial–alveolar CO2 difference can be anything that influences the
efficiency of gas exchange. Specifically, in contrast to the original Bohr dead space calculation, the
physiological dead space calculation will be sensitive to intrapulmonary shunt and diffusion impairment,
and will have a greater sensitivity to alveolar ventilation/perfusion ratio (V′A/Q′) heterogeneity. As the
Enghoff modification of the Bohr dead space measurement is sensitive to a range of physiological gas
exchange abnormalities, it provides a convenient index of overall gas exchange impairment, but it is
important to understand that elevated measurements can arise from multiple mechanisms, dependent on
the specific pathophysiology producing the gas exchange impairment.

DOI: 10.1183/09031936.00137614 1707

PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

Shunt contribution to physiological dead space

Shunt increases not only the alveolar–arterial O2 difference but also the arterial–alveolar CO2 difference
and, therefore, increases calculated physiological dead space. Table 1 presents a simple two-compartment
alveolar model without anatomical dead space, containing a shunt compartment and a normal alveolar
unit. The alveolar space of the shunt compartment is gas that is in equilibrium with the mixed venous gases
and does not contribute to the exhaled gas mixture. One third of the total perfusion of 4.5 L is delivered to
the shunt unit and all of the ventilation is delivered to the normal lung unit, yielding a V′A/Q′ of 0 for the
shunt unit, a V′A/Q′ of 1.5 for the normal unit, and an overall V′A/Q′ for the two units combined of 1.0.
Assuming a mixed venous PCO2 of 46 mmHg, and sufficient alveolar ventilation to reduce the mixed
arterial CO2 to 40 mmHg, the lower alveolar CO2 in the single ventilated compartment produces a 7.5%
contribution to that calculated alveolar dead space (VD/VT−A), even though the model does not contain an
unperfused alveolar unit. While a large shunt is required to achieve this result, note that this
shunt-mediated contribution to the alveolar dead space calculation would be increased by either a decrease
in cardiac output or increase in metabolic rate, where either change would increase the mixed venous PCO2.

V′A/Q′ heterogeneity contribution to physiological dead space

While regional ventilation/perfusion heterogeneity does contribute to the classic Bohr dead space, the
physiological dead space is more heavily influenced by V′A/Q′ heterogeneity. Table 2 provides a simple
three-compartment alveolar model of V′A/Q′ heterogeneity, again without anatomical dead space, with
V′A/Q′ values of 0.1, 1.0 and 10 in the three compartments to illustrate the effect. As with the example in
table 1, the mixed venous CO2 is assumed to be 46 mmHg, and the overall alveolar ventilation is adjusted
so that the mixed PaCO2 will be 40 mmHg. For the three compartments, the overall summed ventilation is
4.2 L and overall summed perfusion is 4.2 L, with an overall V′A/Q′ of 1.0. Combining the CO2 partial
pressures in the ventilated and perfused components of the three units yields a calculated physiological
dead space of 26% in this simple V′A/Q′ heterogeneity model that includes no unperfused alveolar spaces
and no anatomical dead space. While the influence of V′A/Q′ heterogeneity on the alveolar–arterial O2
difference is well known, its influence on CO2 exchange receives little clinical attention, as most patients
can simply increase minute ventilation (V′E) as CO2 exchange becomes less efficient, thereby concealing
the impaired exchange of CO2. Unless a patient requires ventilator support, the requirement for relatively
increased V′E is never noticed. John West employed a more sophisticated computer model of V′A/Q′
heterogeneity to illustrate how disease-relevant degrees of V′A/Q′ heterogeneity would increase
physiological dead space, requiring compensatory increases in V′E to maintain normal PaCO2 values [13].

Physiological dead space in an abnormal lung

In contrast to the simple compartmental models of tables 1 and 2, abnormal lungs ordinarily have
combinations of factors that contribute to the physiological dead space measurement. Figure 2 illustrates a
distribution of V′A/Q′ units typical for an abnormal lung that includes increased ventilation/perfusion
heterogeneity, shunt and anatomical dead space. Note that a normal lung would have the same
bell-shaped distribution of V′A/Q′ units centred on the overall mean V′A/Q′ of 1.0, but the lung units
would be contained within a V′A/Q′ range between 0.5 and 2.0. To quantitatively characterise the relative
roles of combinations of physiological abnormalities on the physiological dead space measurement requires
some familiarity with features of the multiple inert gas elimination technique (MIGET).
To employ MIGET, partial pressures of six intravenously infused inert gases are measured in arterial and
mixed venous blood and mixed expired gas to provide the data required for the mathematical model that

TABLE 1 A simple two-compartment alveolar model containing a shunt compartment and a

normal alveolar unit

Compartment 1 Compartment 2 Combined output

(shunt) (ventilated)

V′A 0.0 L·min−1 4.5 L·min−1 4.5 L·min−1

Q′ 1.5 L·min−1 3.0 L·min−1 4.5 L·min−1
V′A/Q′ 0.0 L·L−1 1.5 L·L−1 1.0 L·L−1
PaCO2 46 mmHg 37 mmHg 40 mmHg
PaCO2 46 mmHg 37 mmHg 37 mmHg
Alveolar VD/VT 7.5%

V′A: alveolar ventilation; Q′: perfusion; PaCO2: arterial CO2 tension; PaCO2: alveolar CO2 tension; VD/VT:
alveolar dead space.

1708 DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

TABLE 2 A simple three-compartment alveolar model of V′A/Q′ heterogeneity

Compartment 1 Compartment 2 Compartment 3 Combined output

(low V′A/Q′) (normal V′A/Q′) (high V′A/Q′)

V′A 0.2 L·min−1 2.0 L·min−1 2.0 L·min−1 4.2 L·min−1

Q′ 2.0 L·min−1 2.0 L·min−1 0.2 L·min−1 4.2 L·min−1
V′A/Q′ 0.1 L·L−1 1 L·L−1 10 L·L−1 1 L·L−1
PaCO2 42 mmHg 40 mmHg 20 mmHg 40 mmHg
PaCO2 42 mmHg 40 mmHg 20 mmHg 29.6 mmHg
Alveolar VD/VT 26%

V′A: alveolar ventilation; Q′: perfusion; PaCO2: arterial CO2 tension; PaCO2: alveolar CO2 tension; VD/VT: alveolar dead space.

describes the distribution of ventilation and perfusion in the lung [14, 15]. The infusion technique with its
associated model provides a quantitative estimate of the allocation of pulmonary blood flow to shunt, to
regions with V′A/Q′ ratios ranging between .001 and 100, and ventilation to inert gas dead space, a
parameter that correlates well with anatomical dead space [16]. While this technique provides a unique
means of fully characterising gas exchange abnormalities in humans, useful insights into the physiological
dead space calculation in disease can be obtained from examining the basic inert gas data itself [17, 18].
In a publication that provided the insights crucial to the subsequent development of MIGET, FARHI [17]
demonstrated that the alveolar or arterial partial pressure of an intravenously infused inert gas could be
predicted based on the solubility of the gas in blood and the V′A/Q′ ratio of the lung unit. For any given
V′A/Q′ value, the predicted arterial (or alveolar) partial pressures of infused inert gases covering a wide
range of solubility in blood form a sigmoid curve when plotted against the log of gas solubilities. The
single solid line in figure 3 illustrates this retention–solubility diagram for a homogenous lung. If lung
units including a distribution of different V′A/Q′ ratios are combined, the arterial and alveolar lines on the
diagram diverge (dashed lines in fig. 3). The vertical line on figure 3 labelled λG represents the solubility
appropriate for CO2, ranging between 2 and 4 mL of gas per mL of blood, depending on the influence of
the Haldane effect [20]. The intersection of the λG line with the arterial and alveolar curves identifies the
two partial pressures needed to make an alveolar dead space calculation for the gas with the solubility λG:

V D=V T ¼ (arterial  alveolar)=arterial

15

10
Frequency %

0
0 0.01 0.1 1.0 10 100
8

V’A/Q’ ratio
Shunt Low V’A/Q’ High V’A/Q’ Dead space

FIGURE 2 Allocation of ventilation and blood flow in an abnormal lung that includes shunt, increased alveolar
ventilation/perfusion ratio (V′A/Q′) heterogeneity and increased anatomical dead space. The lung has an overall V′A/Q′
of 1.0 and has the component lung units sorted according to their individual V′A/Q′ ratios. The broad base of the
bell-shaped curve reflects substantial overall V′A/Q′ heterogeneity. The bar on the left represents the frequency of lung
units compromising shunt, and the bar on the right represents lung units receiving ventilation but no pulmonary artery
blood flow. Figure reproduced courtesy of R.W. Glenny (Division of Pulmonary and Critical Care Medicine, University
of Washington, Seattle, WA, USA).

DOI: 10.1183/09031936.00137614 1709

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

1.0

Pa
0.8
PaHOMO = PAHOMO
Pa/PV or PA/PV

0.6

n
tio
ten PA
Re
0.4

n
io
0.2
et
cr
Ex

0.0
0.001 0.01 0.1 1 λG 10 100 1000
Solubility#

FIGURE 3 Arterial (Pa) (retention) and alveolar (PA) (excretion) partial pressures for intravenously infused inert gases
spanning a very large range of solubility in blood. The single solid curved line represents the arterial and alveolar curves
of a perfectly homogenous lung (PaHOMO = PAHOMO), and the two dashed lines represent the influence of ventilation/
perfusion heterogeneity that creates an arterial–alveolar partial pressure difference for both respiratory and inert gases.
Pv: mixed venous partial pressure; λG: represents the solubility appropriate for CO2. #: solubility is expressed as mL of
gas per mL of blood at 1 Atm. Reproduced from [19] with permission from the publisher.

Applying this graphical inert gas analysis to lung models representing different combinations of abnormal
lung physiology, HLASTALA and ROBERTSON [19] examined the influence of different degrees of V′A/Q′
heterogeneity, shunt and anatomical dead space to illustrate the influence of these abnormalities on a
physiological dead space calculation, examined over a wide range of inert (and respiratory) gas solubility
[19]. Figure 4 illustrates the physiological dead space calculation made over the range of different gas
solubilities in three different abnormal lungs, all containing 20% shunt and 20% anatomical dead space, in
combination with no V′A/Q′ heterogeneity (fig. 4a), normal V′A/Q′ heterogeneity (fig. 4b) and increased
V′A/Q′ heterogeneity (fig. 4c). Note that the physiological dead space in the CO2 solubility range, ∼3 mL
of gas per mL of blood, is most sensitive to the extent of V′A/Q′ heterogeneity, with a secondary sensitivity
to shunt. In the recent context of investigating the mechanisms responsible for dead space abnormalities
in ARDS, WAGNER [21] compared the influence of different degrees of shunt and V′A/Q′ heterogeneity on
physiological dead space, and emphasised the relatively greater influence of V′A/Q′ heterogeneity on the
measurement. A final insight gained from the application of inert gas retention–solubility curves to the
understanding of physiological dead space in disease is that any increase the overall V′A/Q′ ratio (a change
that would shift the bell-shaped distribution in fig. 2 to the right on the V′A/Q′ axis) will also shift all of
the curves on the retention–solubility diagram to the right [17, 18]. Hence a five-fold increase in overall

a) 1.0 b) c)

0.8
Pa/PV or PA/PV

0.6

0.4

0.2

0.0
0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 10 100 0.001 0.01 0.1 1 10 100
Solubility#

FIGURE 4 Plots based on inert gas retention and excretion values for three model lungs that have 20% shunt and 20% anatomical dead space, illustrating the
influence of a) no alveolar ventilation/perfusion ratio (V′A/Q′) heterogeneity, b) normal V′A/Q′ heterogeneity, and c) a high extent of V′A/Q′ heterogeneity. The
dashed lines identify the physiological dead space calculation for the entire range of inert gas solubility. The solid lines represent the inert gas arterial–alveolar
differences and the dotted lines represent the venous admixture calculation for the inert gases. #: solubility is expressed as mL of gas per mL of blood at 1 Atm.
Reproduced from [19] with permission from the publisher.

1710 DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

V′A/Q′ ratio will shift all of the figure 4 curves five units to the right along the solubility axis. As gas
solubility in blood is fixed, any increase in the mean V′A/Q′ value by increased ventilation and/or
decreased perfusion will also increase the calculated physiological dead space. Of note, the influence of a
substantial increase in the mean V′A/Q′ ratio on the physiological dead space measurement was first
described by John West, based on calculations utilising his initial computer model of ventilation/perfusion
interactions in the lung [22].

Physiological dead space during exercise

The initial studies of gas exchange during a progressive work exercise test by WASSERMAN et al. [23]
demonstrated a progressive decrease in physiological dead space from rest to maximal exercise, with a
normal response identified as a VD/VT of <20% at maximal effort. This decrease is readily explained by
the progressive increase in tidal volume during the first half of a progressive work exercise test to ∼60% of
the vital capacity, associated with an anatomical dead space that increases only trivially with the increase
in end-inspiratory volume induced by exercise. The decrease in dead space during exercise was originally
postulated to also include improved ventilation/perfusion matching, but measurements of ventilation/
perfusion heterogeneity during exercise in normal humans assessed by MIGET have not demonstrated a
decrease in overall ventilation/perfusion heterogeneity at low and moderate levels of exercise [24], the
period in a progressive work exercise test when the VD/VT decreases most appreciably [23]. At the highest
tolerated levels of exercise, V′A/Q′ heterogeneity increases modestly [24], and is associated with the
well-known increase in the alveolar–arterial O2 difference. From these MIGET observations, it might be
expected that physiological VD/VT should increase slightly with maximal effort, but increases at maximal
effort are inconsistent and small in comparison with the overall decrease in VD/VT seen early during a
progressive exercise work test.

Can “true” alveolar dead space also exchange CO2?

The original interpretation of alveolar dead space was as a compartment in alveolar spaces that received no
pulmonary artery perfusion, although the current definition of the term alveolar dead space refers to
anything in the physiological dead space that does not include anatomical dead space [12]. While
ventilated alveolar regions without pulmonary artery perfusion (here defined as “true” alveolar dead space)
develop following acute pulmonary emboli and should lead to abnormal dead space measurements,
physiological dead space as a diagnostic test for pulmonary emboli was found to lack sensitivity for all but
the most massive occlusions. Some of that lack of sensitivity may be secondary to gas exchange provided
by the bronchial circulation. Alveolar regions of lung with acute loss of pulmonary artery perfusion can
receive capillary perfusion in variable amounts from connections to the bronchial circulation and drain
into the mixed pulmonary venous output to the left atrium [25]. With sustained occlusion, this bronchial
artery perfusion of alveolar capillaries can be appreciably augmented [25]. Although the systemic bronchial
artery blood perfusing those alveolar capillaries will not be further oxygenated during that second pass
through the lung, it will exchange additional CO2. Hence for segments of embolised lung that only receive
blood from the bronchial arteries, those units function as high V′A/Q′ units from the perspective of CO2
gas exchange, and physiological dead space measurements will underrepresent the true extent of
pulmonary artery occlusion. In the setting of severe bronchiectasis, the elimination of CO2 via the
bronchial circulation can provide a clinically significant contribution to overall CO2 exchange, a situation
occasionally recognised following the embolisation of bronchial arteries for acute haemoptysis in patients
with severe bronchiectasis [26].

Interaction between anatomical dead space and V′A/Q′ heterogeneity

The presence of both anatomical dead space and instrument dead space means that the first portion of
each breath is re-inspired alveolar gas. In the presence of substantial V′A/Q′ heterogeneity, the re-inspired
alveolar gas might be mixed and, therefore, the re-inspired dead space could alter the measured
heterogeneity among the different units of an abnormal lung. The effect of gas re-inspired from a
well-mixed “common” dead space in a lung with overall V′A/Q′ heterogeneity was first analysed utilising
the perspective of the O2–CO2 diagram by ROSS and FARHI [27]. Based on the assumption that the
re-inspired dead space gas was well mixed, they demonstrated that this influence would reduce the
measured extent of overall V′A/Q′ heterogeneity and hence reduce the measured alveolar dead space
contribution to the physiological dead space. A second possibility is that the exhaled dead space gas
remains “personal”, that is, each V′A/Q′ unit would re-inspire its own exhaled gas, an effect that would
have no influence on measured V′A/Q′ heterogeneity. The “common” and “personal” hypotheses for the
distribution of re-inspired gas were modelled by FORTUNE and WAGNER [28] in normal lungs. They
demonstrated that for normal lungs, the assumption that re-inspired dead space gas was “personal” was an
adequate description. However when successive increments of external instrument dead space were added
in an animal study, progressively increasing the proportion of re-inspired gas, PETRINI et al. [29]

DOI: 10.1183/09031936.00137614 1711

PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

demonstrated that there was an appreciable reduction in the extent of externally measured V′A/Q′
heterogeneity. Hence with the addition of larger amounts of instrument dead space, the calculated alveolar
dead space component of the physiological dead space will decrease without changing the actual V′A/Q′
distribution within the lung.

Diffusion impairment and physiological dead space

Carbon dioxide, as a relatively soluble gas in tissue, is not ordinarily considered subject to diffusion
impairment during pulmonary gas exchange. However, the delivery of CO2 to the alveolar spaces from
pulmonary capillary blood requires the catalysis of carbonic anhydrase within the erythrocyte, as the
principal transport of metabolically produced CO2 to the lung is in the form of bicarbonate rather than
dissolved CO2 [13]. Since that chemical reaction requires time, a very fast capillary transit time could
possibly lead to a transpulmonary partial pressure profile identical to that seen with a molecular diffusion
impairment. Model studies have suggested that PCO2 in alveolar gas might be slightly lower that the
systemically sampled PaCO2 if there was decreased alveolar capillary residence time for this reaction [30].
In addition, any drug that impairs the action of erythrocyte carbonic anhydrase can increase the PaCO2 to
PaCO2 difference, a reversible drug effect previously documented in burn patients treated with the old
topical sulfonamides with substantial carbonic anhydrase inhibitory effects [31]. Acetazolamide, a weak
diuretic currently used to prevent symptoms in high altitude sojourners, acts by inhibition of renal
carbonic anhydrase, and also influences erythrocyte carbonic anhydrase. While inhibition could produce a
mild arterial–alveolar CO2 gradient in the lung, at recommended clinical doses a measurable influence
would probably only be apparent during heavy exercise [30].

Dead space measurements in ARDS

Clinical interest in the physiological dead space measurement was reawakened by the publication by NUCKTON
et al. [32] in 2002, linking dead space measurements to prognosis in the ARDS. Physiological dead space was
measured in 179 mechanically ventilated ARDS patients on the day of the syndrome onset. The mean dead
space fraction was 0.54 in eventual survivors and 0.63 in patients who succumbed to the syndrome, and the
risk of death increased with every 0.05 increment in dead space. The physiological dead space measurement
outperformed all of the previous prognostic measures including traditional measures of oxygenation
impairment, lung compliance and illness severity. The authors postulated that the abnormal measurement
was due to regions with blocked microcirculation that remained ventilated, in short, that the abnormality was
secondary to the creation of “true” alveolar dead space. CEPOVKA et al. [33] performed a subsequent study of
ARDS patients ventilated with 6 cm3·kg−1 tidal volumes, demonstrating a nearly identical predictive power of
physiological dead space measurements, but no significant correlation between pulmonary artery pressures
and the dead space measurements. Further confirmation of the prognostic value of physiological dead space
measurement in both acute lung injury (ALI) and ARDS was presented by SIDDIKI et al. [34] utilising
previously collected ARDS network data from 1896 patients. While arterial blood gas and ventilator
measurements were available, the mixed expired CO2 values were not measured, and hence mixed expired
CO2 had to be calculated based on predicted CO2 production rates appropriate for acutely ill mechanically
ventilated patients. Based on the arterial CO2 measurements and predicted mixed expired CO2
concentrations, SIDDIKI et al. [34] found that at both day 1 and day 3 of ARDS diagnosis, patients with a dead
space fraction in excess of 0.50 had a risk for death that increased with every additional 0.10 increment in
dead space fraction, a risk prediction that almost exactly equalled the predictive power of the complete CO2
measurements on ARDS patients described by NUCKTON et al. [32] and CEPOVKA et al. [33].

Factors contributing to the elevated VD/VT in ARDS

By contrast to the postulated development of unperfused regions of lung in ARDS to explain the elevated
physiological dead space, both increased V′A/Q′ heterogeneity and shunt are the more likely contributors
to that observation [21, 35]. However, because the gas exchange characteristics of ARDS are multifactorial,
it is instructive to review previous studies of ALI performed using MIGET, which provides quantitative
measurement of the influence of shunt, V′A/Q′ heterogeneity and dead space. COFFEY et al. [36] utilised an
oleic acid model of ALI to study the influence of positive end-expiratory pressure (PEEP) on dead space
calculations, including MIGET measurements to fully characterise the gas exchange responses. After the
injury stabilised in the anaesthetised animals, gas exchange measurements were acquired at randomly
allocated levels of PEEP ranging between 0 and 20 cmH2O. Physiological dead space measurements were
made at each level of PEEP, and the contributions of the components of shunt, V′A/Q′ heterogeneity,
anatomical dead space, and alveolar dead space (defined as V′A/Q′ >100) were noted. In this animal
model, progressive addition of PEEP improved both shunt and V′A/Q′ heterogeneity, with an initial
reduction in the physiological dead space. However, higher levels of PEEP increased the Fowler dead
space, accounting for increased ventilation to regions with V′A/Q′ >100, and modestly increased the
physiological dead space. In this animal model with randomly allocated PEEP levels, the high V′A/Q′

1712 DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

regions resolved when PEEP levels were lowered, so the development of very high V′A/Q′ regions with
high PEEP was the result of PEEP-created increases in Fowler dead space and zone 3 regions of lung
rather than ventilated lung regions with microvascular blockade that had been postulated in the human
ARDS study of NUCKTON et al. [32].
Three studies utilising MIGET to investigate gas exchange in ARDS patients, all performed in the era when
application of high tidal volumes (10–15 cm3·kg−1) was the clinical standard, described various
combinations of shunt, very low V′A/Q′ regions and increased overall V′A/Q′ heterogeneity, but with an
infrequent incidence of isolated high V′A/Q′ regions among all patients studied [37–39]. The study of RALPH
et al. [38] included data obtained during progressive application of PEEP and did not observe the
development of high V′A/Q′ units at the higher levels of PEEP that had been noted in the dog model of
ALI. A current era MIGET study of ARDS patients by FEIHL et al. [40] comparing the gas exchange
responses to ventilation with tidal volumes of 10 cm3·kg−1 or 6 cm3·kg−1 showed minimal evidence of high
V′A/Q′ regions at 10 cm3·kg−1 tidal volumes. With the 6 cm3·kg−1 tidal volumes, the shunt fraction
increased, but evidence for high V′A/Q′ regions had essentially vanished, as evidenced by the very small
retention–excretion differences for the most soluble gases. The patient population in the FEIHL et al. [40]
study had an average physiological VD/VT of 65% at both 10 cm3·kg−1 and 6 cm3·kg−1 tidal volumes,
and while the lower tidal volume did have a proportionately larger anatomical dead space, the unchanged
VD/VT observed at 6 cm3·kg−1 was explained by the higher cardiac output and lower mixed venous PCO2 at
that lower tidal volume. This ARDS patient population, with no demonstrable high V′A/Q′ component to
their gas exchange when receiving 6 cm3·kg−1 tidal volumes, had a higher overall VD/VT than the mean
values from patients in the NUCKTON et al. [32] and CEPOVKA et al. [33] studies. Hence, MIGET studies of
patients with ARDS have not revealed any consistent elevation in unperfused alveolar dead space, and
indeed the MIGET study performed utilising the current 6 cm3·kg−1 tidal volumes demonstrated a striking
absence of any high V′A/Q′ component, despite severe elevations in the physiological VD/VT measurement.
A final potential influence on the physiological VD/VT in the most severely ill ARDS patients relates to the
overall V′A/Q′ ratio. For patients with a very low cardiac output who are receiving very high V′E, the
overall V′A/Q′ ratio may exceed 5 L·L−1. An increase in the mean V′A/Q′ ratio will shift all the curves on
the retention–solubility diagram to the right, increasing the physiological dead space measurement, just as
a decrease in the mean V′A/Q′ ratio would have the opposite effect. Hence while physiological dead space
abnormalities in ARDS patients are a compelling indicator of prognosis, the interpretation that abnormal
values primarily represent the creation of lung parenchyma that is ventilated but not perfused is not
consistent with the human data obtained utilising MIGET. The physiological dead space abnormalities in
ARDS patients supported with the currently utilised 6 cm3·kg−1 tidal volumes arise primarily as a
consequence of the presence of both shunt and increased low and mid-range V′A/Q′ heterogeneity [40].

Abnormal exercise dead space measurements in heart failure

Exercise studies done on patients with stable severe heart failure have demonstrated an abnormally
elevated exercise ventilation response that is present at all levels of exertion. The ventilatory equivalent for
CO2 (V′E/carbon dioxide production (V′CO2)), calculated as the slope of the two measurements acquired
throughout a maximal progressive exercise test, has become a standard tool for evaluation of patients with
severe heart failure [41]. The finding of a V′E/V′CO2 >34 L·L−1 is associated with adverse cardiac
outcomes, and the risk increases progressively for patients with additional increases in that measurement
[42]. The elevated V′E/V′CO2 in a heart failure patient is primarily attributable to alveolar hyperventilation,
as peripheral chemoreceptor hyperactivity, demonstrated by an enhanced ventilation response to hypoxia,
is a consistent feature in heart failure patients with significant impairment [43]. However, that augmented
exercise ventilation response in the most impaired patients is also associated with an elevated physiological
dead space during exercise [44–46]. The hypothesis that this elevated exercise dead space reflected regions
of very high V′A/Q′ in the lung [45] was not consistent with the single MIGET study of heart failure
patients at rest that revealed no increases in high V′A/Q′ regions or increases in inert gas dead space [47].
While there are currently no MIGET studies of severe heart failure patients during exercise, an alternative
explanation for the increased dead space arises as a consequence of the exercise hyperventilation and
impaired cardiac output. The consistent feature of the exercise response in severe heart failure patients is
the exceptionally high ventilation relative to cardiac output during a progressive work exercise test. The
most impaired patients can barely double their cardiac output from rest to maximal effort [48]. That
disproportionate increase in exercise ventilation associated with an impaired cardiac output can yield
overall V′A/Q′ ratios that may approach 10 L·L−1 during exercise in the most impaired patients. That
increase in overall V′A/Q′ ratio will shift the retention and excretion curves in figure 3 to the right, and
therefore also shift all the derived curves in figure 4 to the right. As the solubility of CO2 remains
unchanged, the exercise associated V′A/Q′ shift leads to an increased physiological dead space
measurement for CO2. Hence a consistent explanation for the elevated dead space measurements seen

DOI: 10.1183/09031936.00137614 1713

PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

during exercise in the most impaired heart failure patients is the abnormally elevated overall V′A/Q′ ratio
and its interaction with V′A/Q′ heterogeneity [49, 50].

Abnormal exercise dead space measurements in pulmonary hypertension

While normal subjects ordinarily demonstrate at least a 50% reduction in physiological dead space during
heavy exercise compared with their resting measurement, patients with pulmonary hypertension typically
fail to show any reduction in dead space during exercise, despite demonstrating typical increases in tidal
volume as the exercise intensity increases. The original explanation proposed for this observation was that
exercise produced more high V′A/Q′ regions in the lungs of those patients. A MIGET study of resting
patients with idiopathic pulmonary hypertension (IPH) or chronic thromboembolic pulmonary
hypertension (CTEPH) by DANTZKER et al. [51] revealed a modest increase in both V′A/Q′ heterogeneity
and shunt, although resting physiological dead space still remained within normal limits. A follow-up
study by DANTZKER et al. [52], which included MIGET measurements taken during exercise on seven
pulmonary hypertension patients, found that the extent of overall V′A/Q′ heterogeneity during exercise
remained unchanged. The exercise study patients had a mean resting physiological VD/VT of 42% that
increased to 44% with maximal effort. The patients showed the characteristic haemodynamic and
ventilatory responses to exercise seen in patients with severe pulmonary hypertension, with marked
exercise hyperventilation and only modest increases in cardiac output leading to exercise mean V′A/Q′
ratios in the range of 3–12. As the extent of V′A/Q′ heterogeneity and shunt fraction remained unchanged
in the transition from rest to exercise, this study represents the best documented instance of the influence
of an exercise-elevated mean V′A/Q′ ratio on the physiological dead space measurement. While the
increase in mean V′A/Q′ with exercise appears to be an important determinant of the abnormal
physiological dead space measurement in pulmonary hypertension patients, their underlying V′A/Q′
heterogeneity and shunt are still important contributors to the exercise measurement abnormality, as
normal subjects at maximal exercise will show at least two-fold increases in mean V′A/Q′, and still
manifest the normal decrease in VD/VT with progressive exercise.
A recent study by ZHAI et al. [53] comparing exercise physiological dead space measurements in patients
with IPH and CTEPH found higher VD/VT in the CTEPH patients, despite comparable resting
haemodynamic measurements in the two groups. In addition, unlike the IPH patients, the functional status
of the CTEPH patients did not correlate with the extent of physiological dead space abnormality, and the
authors suggested that the balance of factors determining the physiological dead space in CTEPH during
exercise might be different from that in IPH. DANTZKER et al. [51] did not identify an increased inert gas
dead space in the three CTEPH patients included in their study, but as the bronchial artery flow increases
dramatically to chronically embolised regions of lung [25], those chronically embolised regions will
function as high V′A/Q′ units, thereby increasing the V′A/Q′ heterogeneity contribution to the physiological
dead space calculation. Another explanation for the higher exercise VD/VT observed in the CTEPH patients
could be a relatively higher mean exercise V′A/Q′ in comparison with the IPH patients [53].

Summary and conclusion

The physiological dead space is defined as including anatomical dead space and alveolar dead space
components. In normal subjects, the measurement is primarily determined by the contribution of the
anatomical dead space, with a small addition from the alveolar dead space attributable to normal
ventilation/perfusion heterogeneity. However, in a wide range of pulmonary disease conditions the alveolar
dead space component becomes more important, and the original concept that it reflects the influence of
regions of lung parenchyma receiving no pulmonary artery perfusion is never an adequate explanation.
Any of the physiological mechanisms contributing to an increased arterial–alveolar CO2 difference will
increase the measured physiological dead space, but ordinarily the extent of overall V′A/Q′ heterogeneity is
the most important contributor. In addition, in the presence of increased V′A/Q′ heterogeneity, any
stimulus increasing the overall mean V′A/Q′, such as extreme hyperventilation with exercise, will further
increase the physiological dead space measurement.

References
1 Bohr C. [Ueber die Lungenathmung]. Skand Arch Physiol 1891; 2: 236–268.
2 Krogh A, Lindhard J. The volume of the dead space in breathing and the mixing of gases in the lungs of man.
J Physiol 1917; 51: 59–90.
3 Jones NL, Robertson DG, Kane JW. Difference between end-tidal and arterial PCO2 in exercise. J Appl Physiol
Respir Environ Exerc Physiol 1979; 47: 954–960.
4 Suarez-Sipmann F, Santos A, Böhm SH, et al. Corrections of Enghoff’s dead space formula for shunt effects still
overestimate Bohr’s dead space. Respir Physiol Neurobiol 2013; 189: 99–105.
5 Fowler WS. Lung function studies; the respiratory dead space. Am J Physiol 1948; 154: 405–416.
6 Klocke RA. Dead space: simplicity to complexity. J Appl Physiol 2006; 100: 1–2.

1714 DOI: 10.1183/09031936.00137614

 PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

7 Bartels J, Severinghaus JW, Forster RE, et al. The respiratory dead space measured by single breath analysis of
oxygen, carbon dioxide, nitrogen or helium. J Clin Invest 1954; 33: 41–48.
8 Crawford AB, Makowska M, Paiva M, et al. Convection- and diffusion-dependent ventilation maldistribution in
normal subjects. J Appl Physiol 1985; 59: 838–846.
9 Engel LA, Menkes H, Wood LD, et al. Gas mixing during breath holding studied by intrapulmonary gas sampling.
J Appl Physiol 1973; 35: 9–17.
10 Fukuchi Y, Roussos CS, Macklem PT, et al. Convection, diffusion and cardiogenic mixing of inspired gas in the
lung; an experimental approach. Respir Physiol 1976; 26: 77–90.
11 Enghoff H. [Volumen inefficax. Bermekungen zur Frage des shadlichen Raumes. Upsala Laekarefoeren]. Foerh
1938; 44: 191–218.
12 West JB. Respiratory Physiology – The Essentials. 6th Edn. Philadelphia, Lippincott, Williams & Wilkins, 2000
13 West JB. Causes of carbon dioxide retention in lung disease. N Engl J Med 1971; 284: 1232–1236.
14 Wagner PD, Naumann PF, Laravuso RB. Simultaneous measurement of eight foreign gases in blood by gas
chromatography. J Appl Physiol 1974; 36: 600–605.
15 Wagner PD, Saltzman HA, West JB. Measurement of continuous distributions of ventilation-perfusion ratios:
theory. J Appl Physiol 1974; 36: 588–599.
16 Dueck R, Wagner PD, West JB. Effects of positive end-expiratory pressure on gas exchange in dogs with normal
and edematous lungs. Anesthesiology 1977; 47: 359–366.
17 Farhi LE. Elimination of inert gas by the lung. Respir Physiol 1967; 3: 1–11.
18 West JB, Wagner PD, Derks CM. Gas exchange in distributions of VA-Q ratios: partial pressure-solubility
diagram. J Appl Physiol 1974; 37: 533–540.
19 Hlastala MP, Robertson HT. Inert gas elimination characteristics of the normal and abnormal lung. J Appl Physiol
Respir Environ Exerc Physiol 1978; 44: 258–266.
20 Robertson HT, Hlastala MP. Elevated alveolar PCO2 relative to predicted values during normal gas exchange.
J Appl Physiol Respir Environ Exerc Physiol 1977; 43: 357–364.
21 Wagner PD. Causes of a high physiological dead space in critically ill patients. Crit Care 2008; 12: 148.
22 West JB. Ventilation-perfusion inequality and overall gas exchange in computer models of the lung. Respir Physiol
1969; 7: 88–110.
23 Wasserman K, Van Kessel AL, Burton GG. Interaction of physiological mechanisms during exercise. J Appl
Physiol 1967; 22: 71–85.
24 Wagner PD, Gale GE, Moon RE, et al. Pulmonary gas exchange in humans exercising at sea level and simulated
altitude. J Appl Physiol 1986; 61: 260–270.
25 Deffebach ME, Charan NB, Lakshminarayan S, et al. The bronchial circulation. Small, but a vital attribute of the
lung. Am Rev Respir Dis 1987; 135: 463–481.
26 Henig NR, Glenny RW, Aitken ML. A hypertrophied bronchial circulatory system may participate in gas
exchange. Lancet 1998; 351: 113.
27 Ross BB, Farhi LE. Dead-space ventilation as a determinant in the ventilation-perfusion concept. J Appl Physiol
1960; 15: 363–371.
28 Fortune JB, Wagner PD. Effects of common dead space on inert gas exchange in mathematical models of the
lung. J Appl Physiol 1979; 47: 896–906.
29 Petrini MF, Robertson HT, Hlastala MP. Interaction of series and parallel dead space in the lung. Respir Physiol
1983; 54: 121–136.
30 Swenson ER, Maren TH. A quantitative analysis of CO2 transport at rest and during maximal exercise. Respir
Physiol 1978; 35: 129–159.
31 White MG, Asch MJ. Acid-base effects of topical mafenide acetate in the burned patient. N Engl J Med 1971; 284:
1281–1286.
32 Nuckton TJ, Alonso JA, Kallet RH, et al. Pulmonary dead-space fraction as a risk factor for death in the acute
respiratory distress syndrome. N Engl J Med 2002; 346: 1281–1286.
33 Cepkova M, Kapur V, Ren X, et al. Pulmonary dead space fraction and pulmonary artery systolic pressure as early
predictors of clinical outcome in acute lung injury. Chest 2007; 132: 836–842.
34 Siddiki H, Kojicic M, Li G, et al. Bedside quantification of dead-space fraction using routine clinical data in
patients with acute lung injury: secondary analysis of two prospective trials. Crit Care 2010; 14: R141.
35 Robertson HT, Swenson ER. What do dead-space measurements tell us about the lung with acute respiratory
distress syndrome?. Respir Care 2004; 49: 1006–1007.
36 Coffey RL, Albert RK, Robertson HT. Mechanisms of physiological dead space response to PEEP after acute oleic
acid lung injury. J Appl Physiol Respir Environ Exerc Physiol 1983; 55: 1550–1557.
37 Dantzker DR, Brook CJ, Dehart P, et al. Ventilation-perfusion distributions in the adult respiratory distress
syndrome. Am Rev Respir Dis 1979; 120: 1039–1052.
38 Ralph DD, Robertson HT, Weaver LJ, et al. Distribution of ventilation and perfusion during positive
end-expiratory pressure in the adult respiratory distress syndrome. Am Rev Respir Dis 1985; 131: 54–60.
39 Mélot C, Lejeune P, Leeman M, et al. Prostaglandin E1 in the adult respiratory distress syndrome. Benefit for
pulmonary hypertension and cost for pulmonary gas exchange. Am Rev Respir Dis 1989; 139: 106–110.
40 Feihl F, Eckert P, Brimioulle S, et al. Permissive hypercapnia impairs pulmonary gas exchange in the acute
respiratory distress syndrome. Am J Respir Crit Care Med 2000; 162: 209–215.
41 Coats AJ. Why ventilatory inefficiency matters in chronic heart failure. Eur Heart J 2005; 26: 426–427.
42 Ponikowski P, Francis DP, Piepoli MF, et al. Enhanced ventilatory response to exercise in patients with chronic
heart failure and preserved exercise tolerance: marker of abnormal cardiorespiratory reflex control and predictor of
poor prognosis. Circulation 2001; 103: 967–972.
43 Ponikowski P, Chua TP, Anker SD, et al. Peripheral chemoreceptor hypersensitivity: an ominous sign in patients
with chronic heart failure. Circulation 2001; 104: 544–549.
44 Myers J, Salleh A, Buchanan N, et al. Ventilatory mechanisms of exercise intolerance in chronic heart failure.
Am Heart J 1992; 124: 710–719.
45 Wasserman K, Zhang YY, Gitt A, et al. Lung function and exercise gas exchange in chronic heart failure.
Circulation 1997; 96: 2221–2227.

DOI: 10.1183/09031936.00137614 1715

PHYSIOLOGY IN RESPIRATORY MEDICINE | H.T. ROBERTSON

46 Johnson RL Jr. Gas exchange efficiency in congestive heart failure II. Circulation 2001; 103: 916–918.
47 Adnot S, Radermacher P, Andrivet P, et al. Effects of sodium-nitroprusside and urapidil on gas exchange and
ventilation-perfusion relationships in patients with congestive heart failure. Eur Respir J 1991; 4: 69–75.
48 Agostoni PG, Wasserman K, Perego GB, et al. Non-invasive measurement of stroke volume during exercise in
heart failure patients. Clin Sci 2000; 98: 545–551.
49 Robertson HT. Gas exchange consequences of left heart failure. Compr Physiol 2011; 1: 621–634.
50 Robertson HT. Excess ventilation during exercise and prognosis in chronic heart failure. Am J Respir Crit Care
Med 2012; 185: 227.
51 Dantzker DR, Bower JS. Mechanisms of gas exchange abnormality in patients with chronic obliterative pulmonary
vascular disease. J Clin Invest 1979; 64: 1050–1055.
52 Dantzker DR, D’Alonzo GE, Bower JS, et al. Pulmonary gas exchange during exercise in patients with chronic
obliterative pulmonary hypertension. Am Rev Respir Dis 1984; 130: 412–416.
53 Zhai Z, Murphy K, Tighe H, et al. Differences in ventilatory inefficiency between pulmonary arterial hypertension
and chronic thromboembolic pulmonary hypertension. Chest 2011; 140: 1284–1291.

1716 DOI: 10.1183/09031936.00137614